ABSTRACT A distinguishing feature of the brain is that its circuitry isn?t computationally static, it adapts to experience. Understanding the circuit mechanisms for adaptive behavior carries two-fold potential benefits - revealing the brain?s learning rules and identifying key behaviorally significant functional ?nodes?. These nodes suggest potent sites to target for therapy development and may also be instructive to suggest more basic circuit principles underlying behavior. Using striatal circuitry and habit learning as a model system, we recently uncovered a set of paradigm- challenging findings in a striatum-dependent habit learning task. In particular, we discovered a new circuit-level signature, termed dviLP (direct vs indirect Latency Plasticity), which distinguishes striatal slices prepared from habitual vs goal-directed animals. The features of dviLP shift long-held attention on rate differences between the two principle projection neuron types, those to the direct and indirect pathways, to consider that behaviorally adaptive signals may be generated by plasticity of their relative timing to fire. Moreover, the origin of this plasticity appears to involve striatal fast-spiking interneurons, a highly non-canonical site for the expression of long-lasting plasticity. Beginning with this highly novel foundation, here we propose to generate a robust predictive computational model for striatal-dependent learning mechanisms by joining multiple disciplines and multiple levels of analysis through an iterative process of circuit modeling and experimentation. In Aim 1, we will comprehensively map functional changes in synaptic and cellular activity that define the behavioral transition from goal-directed to habitual in an operant lever press task. We will use a layered suite of molecular genetic tools to assign coordinates that specify inputs, outputs, compartments (striosome/matrix) and regions (medial, dorsal). In Aim 2, we will measure the activity of genetically specified components of the striatum in behaving mice, identifying the dynamic changes that correlate with and cause the shift from goal- directed to habitual behavior. Our team offers multidisciplinary strengths. Dr. Calakos and Yin have expertise in habit behavior, plasticity mechanisms and in vivo circuit dynamics; ideal for spearheading this effort. The success and impact of this effort will be amplified by tightly incorporating Dr. Brunel?s expertise in computationally modeling brain learning mechanisms and Dr. Tadross?s novel pharmacogenetic reagents that are ideally positioned to test causality of synaptic plasticity events, offering the unique opportunity to manipulate a specific synaptic receptor in a genetically defined cell type. Ultimately, we expect that the knowledge gained through this highly collaborative proposal will provide a foundational resource to accelerate understanding of striatal learning rules for adaptive behavior.